基于3S技术的达里诺尔湖水深反演研究

发布时间：2018-08-26 20:59

【摘要】：水深测量是湖泊调查中的重要工作之一,了解有关区域水深的变化状况对水资源利用、洪涝监测、防灾治灾等方面都至关重要。目前达里诺尔湖的水体监测方法中,水深指标的测量仍然使用传统测深手段,利用遥感技术实现大范围同步测深的研究还未见报道,本文首次尝试运用3S技术,利用3种遥感卫星共6景影像进行对比分析来反演达里诺尔湖深度,构建水下湖底地形。得到了以下结论:(1)可见光波段中,红光波段为对达里诺尔湖水深变化较敏感的波段;不可见光波段中,热红外波段在秋末冬初表现出与湖水深度较好的相关关系,分析原因与秋季上下层湖水的热对流现象有关;在对达里诺尔湖水深反演遥感应用中,使用OLI数据比使用ETM+数据或MS数据更加适合。(2)利用了彩色合成法和改进的归一化水体指数法两种水陆分离方法对5副Landsat卫星影像进行了达里诺尔湖湖面的提取,发现2013年湖面面积比2012年湖面面积有所增加,2014年开始慢慢减少,到2016年,达里诺尔湖湖面积恢复到2012年的面积大小;经计算在2012年至2016年,达里诺尔湖湖面面积约为189km2,年变化范围在1%左右。(3)建立了 152个单因子模型,和12个多因子模型,并进行对比筛选。在单因子模型中,2015年遥感影像作为数据源,以B1*B10作为反演因子的指数模型为最佳反演模型;在多因子模型中,2015年单波段+波段组合形式的多元回归模型为最优。在三种数据源的对比中,反演精度由大到小的顺序为OLIMSETM+,说明在对水深反演的相关研究中并非遥感影像的分辨率越高越好,合适的波段设置会取得精度较好的结果。(4)通过多次对比分析,筛选出2015-10-16多因子模型为最优反演模型,利用该模型对影像中每个像元点进行水深计算,并绘制等水深线图、3D格网图和剖面图,分析了湖底地形变化情况,并计算了达里诺尔湖库容量约为13.1亿立方米。
[Abstract]:Bathymetric survey is one of the most important tasks in lake survey. It is very important to understand the variation of regional water depth in water resources utilization, flood monitoring, disaster prevention and treatment, and so on. At present, in the water monitoring method of Darinole Lake, the traditional bathymetric method is still used in the measurement of the water depth index, and the study of using remote sensing technology to realize the large-scale synchronous sounding has not been reported. This paper first tries to use 3s technology. The depth of Darinole Lake is inversed by comparing and analyzing the six images of three remote sensing satellites, and the underwater lake bottom topography is constructed. The conclusions are as follows: (1) in the visible light band, the red band is the one sensitive to the water depth variation of Darinol Lake, and the thermal infrared band in the invisible wave band shows a good correlation with the depth of the lake water at the end of autumn and the beginning of winter. The causes are related to the phenomenon of thermal convection in the upper and lower layers of lake water in autumn, and in the application of remote sensing in inversion of water depth of Darinole Lake, Using OLI data is more suitable than using ETM data or MS data. (2) five sets of Landsat satellite images are extracted from Lake Darinole Lake by using color synthesis method and improved normalized water body index method. The lake area was found to have increased in 2013 over that in 2012, and began to decrease slowly in 2014, before returning to its 2012 size by 2016; calculated from 2012 to 2016, The lake area of Darinole Lake is about 189km2, and the annual variation range is about 1%. (3) 152 single factor models and 12 multi-factor models have been established and compared with each other. In the single factor model, the 2015 remote sensing image is used as the data source, the index model with B1*B10 as the inversion factor is used as the best inversion model, and the multivariate regression model with single band combination in 2015 is the best in the multi-factor model. In the comparison of three data sources, the order of inversion accuracy from large to small is OLIMSETM, which shows that the higher the resolution of remote sensing image, the better the resolution of the remote sensing image in the related research of depth inversion. The appropriate band setting can get better results. (4) by comparing and analyzing several times, the 2015-10-16 multi-factor model is selected as the optimal inversion model, and the depth of each pixel in the image is calculated by using this model. At the same time, the 3D grid map and section map are drawn, the topographic variation of the lake bottom is analyzed, and the capacity of Darinole Lake reservoir is calculated to be about 1.31 billion cubic meters.
【学位授予单位】：内蒙古农业大学
【学位级别】：硕士
【学位授予年份】：2017
【分类号】：P332

【相似文献】